声呐图像背景区域灰度统计特性分析与拟合

发布时间：2018-10-10 07:54

【摘要】：利用声呐进行水下目标定位识别是当前水下目标识别与跟踪的重要手段之一,由于声呐图像受噪声影响严重、分辨率低,对声呐图像的背景建模有助于其目标分割与识别。首先,分析声呐图像背景区域灰度的统计特性,结合其特点采用高斯分布、Gamma分布、威布尔分布、瑞利分布模型对6类不同背景区域声呐图像统计特性进行拟合,构建声呐图像背景区域模型。最后,采用?2准则和Kolmogorov距离误差评价准则评估拟合效果。拟合结果表明,高斯分布、Gamma分布和威布尔分布均能较好地逼近声呐图像背景区灰度统计特性。为满足实时性的应用需求,选用高斯分布构建声呐图像背景灰度统计模型是可行、合理的方案,从而为声呐图像预处理和目标分割提供了背景模型建模的理论依据。
[Abstract]:Underwater target location and recognition using sonar is one of the important methods of underwater target recognition and tracking. Because sonar image is seriously affected by noise and has low resolution, the background modeling of sonar image is helpful to its target segmentation and recognition. Firstly, the statistical characteristics of sonar background region are analyzed, and the statistical characteristics of sonar images are fitted by Gao Si distribution Gamma distribution, Weibull distribution and Rayleigh distribution model. The background region model of sonar image is constructed. Finally, adopt? The fitting effect is evaluated by 2 criteria and Kolmogorov distance error evaluation criterion. The fitting results show that both the Gamma distribution and Weibull distribution of Gao Si distribution can approach the statistical characteristics of the background region of sonar image. In order to meet the requirement of real-time application, it is feasible and reasonable to select Gao Si distribution to construct the background gray scale statistical model of sonar image, which provides the theoretical basis for the background model modeling for sonar image preprocessing and target segmentation.
【作者单位】：三峡大学水电工程智能视觉监测湖北省重点实验室;三峡大学计算机与信息学院;
【基金】：国家自然科学基金(联合基金)重点项目(U1401252);国家自然科学基金资助项目(61272237) 湖北省重点实验室开放基金项目(2015KLA05)
【分类号】：TB56;TP391.41

【相似文献】